适合无线传感器网络的自适应动态电压调节算法

Adaptive dynamic voltage scaling algorithm for wireless sensor networks

对于能量有限的传感器网络,在计算复杂度较高的应用中,节省CPU的能耗具有重要意义。针对以事件为驱动的无线传感器网络的任务模式,提出一种基于零散任务模型的自适应DVS算法——A-DVS。A-DVS算法根据CPU的任务量实时调整工作频率和电压,能在很大程度上降低CPU能耗的同时,保证任务的实时性要求。理论分析和实验结果表明,A-DVS算法的实际节能效果接近理论分析值的80%左右,可在很大程度上延长节点的生命周期。

It is significant to lower energy consumption from CPU for applications with high computing cost in energy-constrained wireless sensor networks. This paper proposed a scattered task model based adaptive DVS （ dynamic voltage scaling） algorithm, A-DVS, in the presence of the task characteristic in event-driven wireless sensor networks. A-DVS algorithm could scale the working frequency and voltage dynamically according to the workload. As a result, it could decrease the CPU energy consumption meanwhile meeting the real-time requirements of tasks. Theoretical and experimental results show that the A-DVS algorithm can achieve 80% or so theoretical energy savings so that prolong the lifetime of sensor nodes.